FHL1C induces apoptosis in notch1-dependent T-ALL cells through an interaction with RBP-J

Aberrantly activated Notch signaling has been found in more than 50% of patients with T-cell acute lymphoblastic leukemia (T-ALL). Current strategies that employ γ-secretase inhibitors (GSIs) to target Notch activation have not been successful. Many limitations, such as non-Notch specificity, dose-limiting gastrointestinal toxicity and GSI resistance, have prompted an urgent need for more effective Notch signaling inhibitors for T-ALL treatment.

R E S E A R C H A R T I C L E Open Access

FHL1C induces apoptosis in notch1-dependent T-ALL cells through an interaction with RBP-J

Wei Fu1†, Kai Wang1†, Jun-Long Zhao2†, Heng-Chao Yu2, San-Zhong Li2, Yan Lin1, Liang Liang2, Si-Yong Huang1, Ying-Min Liang1*, Hua Han2*and Hong-Yan Qin2*

Abstract

Background: Aberrantly activated Notch signaling has been found in more than 50% of patients with T-cell acute lymphoblastic leukemia (T-ALL) Current strategies that employγ-secretase inhibitors (GSIs) to target Notch activa-tion have not been successful Many limitaactiva-tions, such as non-Notch specificity, dose-limiting gastrointestinal toxicity and GSI resistance, have prompted an urgent need for more effective Notch signaling inhibitors for T-ALL treatment Human four-and-a-half LIM domain protein 1C (FHL1C) (KyoT2 in mice) has been demonstrated to suppress Notch activation in vitro, suggesting that FHL1C may be new candidate target in T-ALL therapy However, the role of FHL1C

in T-ALL cells remained unclear

Methods: Using RT-PCR, we amplified full-length human FHL1C, and constructed full-length and various truncated forms of FHL1C Using cell transfection, flow cytometry, transmission electron microscope, real-time RT-PCR, and Western blotting, we found that overexpression of FHL1C induced apoptosis of Jurkat cells By using a reporter assay and Annexin-V staining, the minimal functional sequence of FHL1C inhibiting RBP-J-mediated Notch transactivation and inducing cell apoptosis was identified Using real-time PCR and Western blotting, we explored the possible molecular mechanism of FHL1C-induced apoptosis All data were statistically analyzed with the SPSS version 12.0 software

Results: In Jurkat cells derived from a Notch1-associated T-ALL cell line insensitive to GSI treatment, we observed that overexpression of FHL1C, which is down-regulated in T-ALL patients, strongly induced apoptosis Furthermore,

we verified that FHL1C-induced apoptosis depended on the RBP-J-binding motif at the C-terminus of FHL1C Using various truncated forms of FHL1C, we found that the RBP-J-binding motif of FHL1C had almost the same effect as full-length FHL1C on the induction of apoptosis, suggesting that the minimal functional sequence in the RBP-J-binding motif of FHL1C might be a new drug candidate for T-ALL treatment We also explored the molecular mechanism of FHL1C overexpression-induced apoptosis, which suppressed downstream target genes such as Hes1 and c-Myc and key signaling pathways such as PI3K/AKT and NF-κB of Notch signaling involved in T-ALL progression

Conclusions: Our study has revealed that FHL1C overexpression induces Jurkat cell apoptosis This finding may provide new insights in designing new Notch inhibitors based on FHL1C to treat T-ALL

Keywords: T-cell acute lymphoblastic leukemia, Notch signaling, FHL1C, RBP-J, Apoptosis

* Correspondence: liangym@fmmu.edu.cn ; huahan@fmmu.edu.cn ;

hongyanqinfm@gmail.com

†Equal contributors

1

Department of Hematology, Tangdu Hospital, Fourth Military Medical

University, Xi ’an 710038, People’s Republic of China

2

State Key Laboratory of Cancer Biology, Department of Medical Genetics

and Developmental Biology, Fourth Military Medical University, Xi ’an 710032,

People ’s Republic of China

© 2014 Fu et al.; licensee BioMed Central Ltd This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly credited The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article,

T-cell acute lymphoblastic leukemia (T-ALL) is an

aggres-sive neoplasm that originates from immature T-cells

Although the currently used multi-agents chemotherapy

results in 5-year relapse-free survival rates of over 75%

in children and over 50% in adults, relapse usually is

associated with resistances against chemotherapy and a

very poor prognosis [1-3] Therefore, it is essential to

elucidate the molecular mechanisms underlying T-ALL

progression to discover new therapeutic targets for the

treatment of T-ALL

Mutations in the Notch1 receptor have been

demon-strated as the etiological cause of T-ALL [4,5] The first

evidence of oncogenic Notch signaling was observed in

T-ALL patients, involving translocation of a portion of

the human Notch1 gene to the TCR locus [6] However,

this event is rare in human T-ALL (less than 1%) In fact,

more than 50% of T-ALL patients carry Notch1-activating

mutations that are usually in the heterodimerization (HD)

domain and proline/glutamic acid/serine/threonine-rich

motifs (PEST) of the Notch1 receptor, which result in

delayed degradation of Notch1 [7] Notch1 is one of

the four mammalian Notch receptors that are

single-pass transmembrane proteins consisting of functional

extracellular, transmembrane, and intracellular domains

When the Notch receptor is triggered upon interaction

with its ligands on neighboring cells, the Notch

intracellu-lar domain (NIC) is released from the membrane after

proteolytic cleavages executed by γ-secretase-containing

protease complexes The NIC enters the nucleus and

asso-ciates with the DNA-binding transcription factor RBP-J

through its N-terminal RAM (RBP-J association molecule)

domain, which transactivates promoters harboring

RBP-J-binding sites by dissociating co-repressors, such as

SMRT/N-CoR, HDAC, and MINT [1,8], and recruiting

co-activators including Mastermind-like (MAML) and

p300/CBP [9] In T-ALL, activated Notch1 regulates cell

proliferation and apoptosis by modulating the level and

activities of the related molecules/pathways such as

Hes1, c-Myc, PI3K/AKT, and NF-κB through canonical

(RBP-J-dependent) and/or non-canonical

(RBP-J-inde-pendent) signals [10,11]

Considering the critical role of Notch activation in the

progression of T-ALL, efforts have been made to cure

T-ALL by blocking Notch signaling Small molecule

γ-secretase inhibitors (GSIs), which block the critical

proteolytic steps required for Notch activation, can be

applied for T-ALL treatment, but the clinical outcomes

have been unsatisfactory These outcomes might be

attributed to the fact thatγ-secretase is not specific for

Notch receptors, and more importantly, GSIs only affect

ligand-dependent Notch activation, not ligand-independent

Notch activation resulting from chromosome

transloca-tion or point mutatransloca-tions In additransloca-tion, gastrointestinal

toxicity and weak anti-leukemic effects on T-ALL also hinder the clinical application of GSIs [12,13] Another target for blocking Notch signaling in malignant T cell leukemia is RBP-J that mediates the effects of Notch1 mutants on downstream gene expression Expression of

a dominant-negative MAML1 (DN-MAML1) in T-ALL cell lines has been shown to antagonize Notch1 activa-tion [14,15] Subsequently, Moellering et al designed a stable α-helical peptide derived from MAML1 (SAHM1) based on the structure of DN-MAML1 They found that SAHM1 directly impedes assembly of the Notch1 transac-tivation complex in the nucleus and reduces malignant cell proliferation and promotes apoptosis In contrast to GSIs, DN-MAML1 and SAHM1 inhibit Notch activation more efficiently because of their direct inhibition of Notch signals at the transcriptional factor level However, as a multifunctional transcription activator, MAML1 is also not specific for Notch signaling [16] Thus, more effect-ive Notch signal inhibitors are still required for the treatment of T-ALL

Human four-and-a-half LIM domain protein 1C (FHL1C) (KyoT2 in mice) belongs to the four-and-a-half LIM domain protein family and is an alternatively spliced form of FHL1A/KyoT1 Selective use of exons results in a frame shift in translation, generating a WW-containing motif at the C-terminus of FHL1C, which can bind

to RBP-J Without a transcription activation domain, FHL1C/KyoT2 has been demonstrated to compete with NIC for RBP-J binding and suppress RBP-J-mediated Notch activation in vitro [8] These findings suggest that FHL1C may be another therapeutic target of T-ALL, but the role of FHL1C remains to be investigated in T-ALL cells In the present study, we addressed this issue using T-ALL clinical samples and the T-ALL cell line Jurkat We found that the expression level of FHL1C was lower in the peripheral blood mononuclear cells (PBMCs) of T-ALL patients than that in the controls Overexpression of FHL1C or its various truncates containing the RBP-J-binding site or the minimal RBP-J-RBP-J-binding motif, all resulted in Jurkat cell apoptosis Mechanistically, FHL1C-induced Jurkat cell apoptosis involved suppression of downstream target genes and key pathways of Notch signaling in T-ALL, including PI3K-AKT and NF-κB These findings shed light on the design of new Notch inhibitors based on FHL1C to treat T-ALL

Methods

Vector construction

Total RNA was extracted from a human skeletal muscle biopsy and then reverse transcribed using a commer-cially available kit from TAKARA (Dalian, China) with

an oligo-dT primer This patient had signed informed consent, and the protocol involving human samples was approved by the Ethics Committee of Tangdu Hospital,

Fourth Military Medical University FHL1C (GeneBank

accession number: AF220153.1) was amplified by PCR

with specific primers (Forward primer, 5′-ATGGCGGA

GAAGTTTGACTGCCACTACT-3′; Reverse primer,

5′-TCACGGAGCATTTTTTGCAGTGGAAGCA-3′)

(Additional file 1: Table S1) The 585 bp PCR product

was cloned and confirmed by DNA sequencing The

full-length FHL1C cDNA was inserted into the

expres-sion vectors pEGFP-C1 (Clontech, Mountain View, CA)

and pCMV-Myc (Clontech) to generate pEGFP-FHL1C

and pCMV-Myc-FHL1C, respectively

To construct EGFP-tagged truncates of FHL1C, LIM1,

LIM2, and the C-terminal RBP-J-binding motif (RBPmotif)

of FHL1C, various fragments were subcloned by PCR

with the primers listed in Additional file 1: Table S1, and

pEGFP-FHL1C expression vector was used as the

tem-plate The LIM1 and LIM2 domains were fused in frame

at the 3′ terminus to the RBPmotif

to generate LIM1R and LIM2R, respectively LIM1R, LIM2R, and RBPmotif

were then inserted in frame into pEGFP-C1 to generate

pEGFP-LIM1R, pEGFP-LIM2R, and pEGFP-RBPmotif

(Additional file 2: Figure S3A) To construct vectors for

expression of EGFP fused to the minimal RBPmotif of

FHL1C, double-stranded oligonucleotides encoding

VWWPM, PVWWPMK, and APVWWPMKD peptides

were synthesized and cloned in frame downstream of

EGFP in pEGFP-C1 The plasmids were confirmed by

DNA sequencing

Patients, RNA extraction, RT-PCR, Sequencing

Blood samples were collected from T-ALL patients and

normal healthy individuals (Additional file 3: Tables S3

and Additional file 4: Table S4) All patients and normal

individuals involved in the study had signed informed

consents for the use of their blood samples, except for

children under the age of 18, who had their informed

consents signed by their parents as their representatives

The protocols involving human samples were approved

by the Ethics Committee of Tangdu Hospital, Fourth

Military Medical University Diagnoses had been made

according to standard morphological, immunological,

and molecular genetics criteria PBMCs were separated

by Ficoll-Hypaque density gradient centrifugation Total

RNA was extracted from PBMCs and Jurkat cells using

Trizol reagent (Invitrogen, Carlsbad, CA), and then

re-verse transcribed using the commercially available kit

with random primers cDNA was diluted appropriately

and used for PCR, GAPDH was used as an internal

con-trol DNA sequences corresponding to the HD and

PEST domains were amplified using nested PCR

accord-ing to previous report [7], and then sequencaccord-ing was

per-formed by Biotechnology Company

Real-time PCR was performed as triplicate using

SYBR Premix EX Taq (TAKARA) with an ABI PRISM

7300 real-time PCR system (Applied Biosystems, Life Technologies, Carlsbad, CA) with β-actin as the refer-ence control Primers used for quantitative RT-PCR are listed in Additional file 5: Table S2

Cell culture and transfection

Jurkat cells (ATCC, Rockville, MD) were grown in RPMI

1640 supplemented with 10% fetal calf serum, 2 mM L-glutamate, 100 U/ml penicillin, and 100μg/ml strepto-mycin at 37°C in saturated humidity with 5% CO2 HeLa and Cos7 cells (ATCC) were maintained in Dulbecco’s modified Eagle medium (DMEM) containing the supple-ments mentioned above

HeLa and Cos7 cells were transfected using Lipofecta-mine 2000 (Invitrogen) according to the recommended protocol Jurkat cells (1 × 106) were transfected with a Nucleofector Kit V (Amaxa-Lonza, Cologne, Germany) using a Nucleofector I (program X-01) following the manufacturer’s optimized protocol

Reporter assays

HeLa or Cos7 cells were cultured in 24-well plates and transfected with 5 ng phRL-TK (Promega, Madison, WI), 80 ng pGa981-6 reporter plasmid [17], 200 ng pEF-BOS-Myc-NIC, and serial amounts (100, 300, and

500 ng) of plasmids carrying FHL1C or various truncates

of FHL1C The cells were harvested at 48 h post-transfection, and cell extracts were assayed for luciferase activity using a Gloma X™ 20/20 Luminometer (Promega) The luciferase activity was normalized to Renilla luciferase activity

Flow cytometric analyses of cell cycle progression and apoptosis

Jurkat cells were resuspended in PBS and fixed in 70% ethanol on ice for 2 h The cells were then stained with

20 mg/ml propidium iodide (PI) in PBS containing 0.1% Triton X-100 and 0.2 mg/ml RNase A for 30 min on ice The cells were analyzed by a FACSCalibur flow cyt-ometer (BD Immunocytometry Systems, San Jose, CA) Data were analyzed with CellQuest software

Cell viability was routinely detected by trypan blue exclusion Apoptosis was determined by staining with Annexin V-APC (eBiosciences, San Diego, CA) according

to the manufacturer’s protocol, followed by flow cytomet-ric analysis

Co-immunoprecipitation and western blotting

pEGFP-FHL1C and pCMV-Myc-RBP-J were transfected into HeLa cells Co-immunoprecipitation was performed

as described previously [17] with an anti-Myc antibody (9E10; Santa Cruz Biotechnology, Santa Cruz, CA) Western blotting was performed with anti-FHL1 (ProteinTech, Wuhan, China) or anti-Myc antibodies

Western blotting analysis was performed routinely with

primary antibodies including anti-AKT, anti-phospho-AKT

(Signalway Antibody, Pearland, TX), anti-p50 (3354R-100;

BioVision, Mountain View, CA), or anti-β-actin

(Sigma-Aldrich, St Louis, MO, USA) Anti-rabbit IgG and

anti-mouse IgG (Boster BioTec, Shanghai, China) were used as

secondary antibodies Anti-c-Rel, anti-IκBα antibodies

were purchased from Eptiomics (Abcam, Burlingame,

CA) An caspase 3 antibody (H-277), GFP

anti-body, normal goat IgG, and normal rabbit IgG were

pur-chased from Santa Cruz Biotechnology

Fractionation of subcellular components

Jurkat cells were washed twice with PBS at 4°C and then

resuspended and incubated in buffer A (10 mM Hepes,

1.5 mM MgCl2, 10 mM KCl, 0.5 mM DTT, and a

prote-ase inhibitor cocktail) for 30 min on ice After

centrifu-gation at 4000 rpm for 20 min at 4°C, cytosolic fractions

were collected, and the pellets were washed once in

buf-fer A, resuspended in 1% NP-40 lysis bufbuf-fer (10 mM

Tris–HCl, pH 7.8, 0.5 mM EDTA, 250 mM NaCl, and

the protease inhibitor cocktail), and then incubated for

an additional 30 min on ice After centrifugation at

10000 rpm for 15 min at 4°C, the nuclear factions were

collected Equal amounts of each fraction were analyzed

by SDS-PAGE, followed by western blotting with the

ap-propriate antibodies

Hoechst staining

Cells were washed twice with PBS, fixed in 70% ethanol

for 20 min, and then washed again with PBS Hoechst

diluted at 1:10,000 (final concentration: 0.12 μg/ml)

was added to cells followed by incubation in the dark

for 15 min The cells were washed with PBS and

visu-alized under a fluorescence microscope (BX51; Olympus,

Tokyo, Japan)

Transmission electron microscopy (TEM)

Sample preparation and observation under a transmis-sion electron microscope were performed as described previously [18]

Statistical analysis

Data were analyzed with SPSS version 12.0 software Results were expressed as the mean ± SD Comparisons between groups were performed with the unpaired Student’s t-test A P-value of less than 0.05 was considered statisti-cally significant

Results

FHL1C is down-regulated in PBMCs from T-ALL patients

FHL1C/KyoT2 has been shown to be a negative regula-tor of the Notch pathway by competing with NIC for binding to RBP-J in vitro To assess the relevance of FHL1C in T-ALL, we examined FHL1C mRNA expres-sion in PBMCs from eight T-ALL patients and nine healthy donors as controls by RT-PCR We found that FHL1C mRNA expression was significantly lower in PBMCs from T-ALL patients compared with that in PBMCs from healthy individuals (P < 0.05) (Figure 1A, upper panel and B) Because Hes1 is the main down-stream target gene of activated Notch signaling in T-ALL [19], we also detected Hes1 mRNA expression

in T-ALL and healthy individuals The result showed that Hes1 mRNA expression was significantly higher in T-ALL samples than that in healthy individuals sam-ples (Figure 1A, lower panel and C) These results indi-cate that FHL1C expression is down-regulated in the PBMCs of T-ALL patients

Overexpression of FHL1C induces apoptosis of T-ALL cells

To examine the role of FHL1C in T-ALL, we transiently overexpressed FHL1C in Jurkat cells, a human T-ALL cell line bearing Notch1 activation mutations FHL1C

Figure 1 Expression of FHL1C and Hes1 was detected in T-ALL patients and healthy donors (A) RT-PCR analysis of FHL1C and Hes1 mRNA

in PBMCs from 8 T-ALL patients and 9 healthy donors, with GAPDH as an internal control (B, C) Relative mRNA levels of FHL1C (B) and Hes1 (C) to GAPDH in PBMCs from T-ALL patients and controls were compared The horizontal lines indicated median expression levels.

was fused to EGFP at the N-terminus and introduced

into Jurkat cells by electroporation As determined by

flow cytometric and western blotting analyses, EGFP

expression showed that highly efficient transfection was

achieved in both empty vector and

pEGFP-FHL1C-transfected Jurkat cells (Additional file 6: Figure S1A and

S1B) We monitored cell growth after transfection and

found that the number of EGFP+ Jurkat cells transfected

with pEGFP increased steadily, whereas the number of

EGFP+ Jurkat cells transfected with pEGFP-FHL1C did

not increase significantly and decreased gradually at 36 h

post-transfection (Figure 2A) This observation suggested

that overexpression of FHL1C caused cell growth arrest

and/or cell death in Jurkat cells

We first examined the cell cycle progression of Jurkat

cells transfected with pEGFP or pEGFP-FHL1C The

results showed no remarkable difference in the cell cycle

distribution between the two groups, although the

num-ber of cells overexpressing FHL1C exhibited a slight

increase in G2/M phase (Figure 2B and C) We next

determined cell viability after transfection We found

that the percentage of viable cells decreased

continu-ously among Jurkat cells after transfection with

pEGFP-FHL1C, suggesting that overexpression of FHL1C might

result in cell death (Figure 2D)

Next, we directly estimated apoptosis after

overexpres-sion of FHL1C Jurkat cells were transfected as described

above, and apoptosis was determined by flow cytometric

analysis with annexin-V and PI staining In the GFP+

(transfected) cell population, there was a significant

increase of annexin-V+cells among the

pEGFP-FHL1C-transfected Jurkat cells compared with that among

the pEGFP-transfected Jurkat cells, suggesting that overexpression of FHL1C induced apoptosis in Jurkat cells (Figure 3A and B) Annexin-V and PI staining distin-guishes early apoptotic (annexin V+PI−) and late apop-totic (annexin V+PI+) cells As Figure 3C and D were shown, overexpression of FHL1C resulted in an in-crease of both early and late apoptotic cells among Jurkat cells We also examined the morphology of Jurkat cells transfected with pEGFP or pEGFP-FHL1C by Hoechst staining (Figure 3E) and TEM (Figure 3F) The results confirmed that there were more apoptotic cells with condensed nuclei among Jurkat cells overexpress-ing FHL1C (Figure 3E and F) At the molecular level, overexpression of FHL1C in Jurkat cells reduced the expression of anti-apoptosis molecules, including Bcl-2 and Bcl-x1, and increased expression of the apoptosis-related molecule caspase 3 (Figure 3G and H) These results strongly suggest that overexpression of FHL1C induces apoptosis of T-ALL cells

FHL1C induces apoptosis of Jurkat cells through suppression of RBP-J-mediated transactivation

Similar to its murine homolog KyoT2, FHL1C also possesses

a C-terminal RBPmotif, suggesting that FHL1C interacts with RBP-J and suppresses RBP-J-mediated transactivation

To confirm an interaction between FHL1C and RBP-J,

we performed co-immunoprecipitation HeLa cells were co-transfected with expression vectors for Myc-tagged RBP-J (pCMV-Myc-RBP-J) and EGFP-tagged FHL1C (pEGFP-FHL1C), and immunoprecipitation was per-formed with an anti-Myc antibody Co-precipitated proteins were detected using an anti-FHL1 antibody by

Figure 2 Overexpression of FHL1C in Jurkat cells resulted in reduced cell viability (A) Jurkat cells (5 × 10 6 ) were transfected with pEGFP or pEGFP-FHL1C by using the Nucleofection method The numbers of viable EGFP-positive cells were determined every 12 h by cell counting and FACS analysis (B, C) Cell cycle progression of Jurkat cells in (A) was determined 48 h post-transfection by FACS after PI staining (B) Cells in each phase were compared between the two groups (C) (D) Total viability of Jurkat cells transfected in (A) was monitored by using trypan blue exclusion assay Bars = means ± S.D (n = 3), *P < 0.05, **P < 0.01.

Figure 3 FHL1C overexpression induced apoptosis in Jurkat cells (A) Jurkat cells were transiently transfected with pEGFP or pEGFP-FHL1C

by using the Nucleofection method Apoptosis in the GFP−and GFP + fractions of cells was determined by AnnexinV staining followed by FACS

48 h post-transfection (B) Percentages of apoptotic cells (Annexin V + ) in GFP−and GFP + cell fractions in (A) were compared (C) Jurkat cells were transfected with pEGFP or pEGFP-FHL1C by using the Nucleofection method Early and late apoptotic cells were depicted 48 h post-transfection

by using Annexin V and PI staining followed by FACS (D) GFP + cells in early and late apoptotic phases in (C) were compared (E) Jurkat cells were transiently transfected with pEGFP or pEGFP-FHL1C by using the Nucleofection method Cells were stained with Hoechst 24 h post-transfection and nuclei were observed under a fluorescence microscope Arrow heads indicate Hoechst-positive apoptotic nuclei (F) Typical cell apoptosis in (E) was depicted under TEM Intact cell membrane, organelles and normal nuclear morphology were observed in vector-transfected cells, whereas incomplete membrane and condensed nuclei were observed in cells overexpressing FHL1C (magnification, × 9900) (G) Total RNA was prepared from cells in (E) 24 h post-transfection The mRNA levels of the apoptosis-related molecules were determined by real time RT-PCR, with β-actin as

a reference (H) Cell lysates were prepared from cells in (E) 24 h post-transfection The level of Caspase3 was determined by Western blot analysis Bars = means ± S.D (n = 3), *P < 0.05; NS, not significant.

western blotting analysis The results showed that

GFP-FHL1C was well co-precipitated with RBP-J (Additional

file 7: Figure S2A), suggesting that FHL1C interacts

with RBP-J Furthermore, we performed reporter assays

using HeLa and Cos7 cells by transfection with

pEGFP-FHL1C and a NIC expression vector As a result,

over-expression of FHL1C suppressed transactivation of the

reporter harboring RBP-J-binding sites by NIC in a

dose-dependent manner (Additional file 7: Figure S2B)

This result demonstrated that FHL1C suppresses

RBP-J-mediated transactivation by competing with NIC

We next determined whether FHL1C induced

apop-tosis of Jurkat cells through suppression of

RBP-J-mediated transactivation by overexpressing RBP-J-VP16,

a constitutively activated RBP-J [20] Jurkat cells were

transfected with pEGFP-FHL1C alone or co-transfected

with pEGFP-FHL1C and pCMX-VP16-RBP-J, followed

by analysis of apoptosis The results showed that Jurkat

cells did not undergo apoptosis after transfection with

pCMX-VP16-RBP-J alone, and overexpression of FHL1C

alone induced apoptosis, which was consistent with the

results shown above Co-transfection of cells with

vec-tors carrying FHL1C and RBP-J-VP16 resulted in

effi-cient attenuation of the FHL1C-induced apoptosis

(Figure 4A) This effect was proportional to the amount

of RBP-J-VP16 (Figure 4B) These data suggest that

con-stitutively activated RBP-J protects Jurkat cells from

FHL1C-induced apoptosis, most likely through

constitu-tive activation of Notch target genes

The C-terminal RBPmotifof FHL1C is sufficient to induce

apoptosis of Jurkat cells

FHL1C/KyoT2 is composed of two N-terminal LIM

do-mains and a 27 amino acid RBPmotif at the C-terminus

[21] To determine which domain of FHL1C is critical

for FHL1C-induced apoptosis of Jurkat cells, various

EGFP fusion proteins in which EGFP was fused to

full-length FHL1C, LIM1R, LIM2R, or RBPmotif were

trans-fected into HeLa cells and then visualized under a

confocal fluorescence microscope As a result, these

fu-sion proteins showed similar subcellular localization

(Additional file 2: Figure S3A and S3B) Next, we examined

the effect of these fusion proteins on RBP-J-mediated

trans-activation using a reporter assay The results showed that

all of the fusion proteins exhibited a transcription

suppres-sion effect on RBP-J-mediated transactivation of the

re-porter gene (Additional file 2: Figure S3C), although the

full-length FHL1C fusion protein had the strongest activity

We next evaluated the ability of these fusion proteins

to induce apoptosis of Jurkat cells Jurkat cells were

transfected with each of the constructs, and apoptosis

was assessed at 24 h post-transfection We found that

transfection of each construct induced apoptosis of Jurkat

cells (Figure 5A) The number of GFP+ cells decreased

continuously after transfection, except for EGFP-LIM1R-overexpressing cells that showed a decrease in cell number before 36 h post-transfection followed by an increase in the number of GFP+cells (Figure 5B) We next examined the mRNA expression of critical downstream genes of Notch signaling, which are involved in T-ALL cells includ-ing Hes1 [19], Pten [22,23], p53 [24], and c-Myc [25,26], and apoptosis-related genes Bcl2, BAX [27], and caspase 3 [28] The results showed that all of the fusion proteins down-regulated the expression of Hes1 and c-Myc, but EGFP-LIM1R only showed a mild effect Consistent with the FHL1C-induced apoptosis, overexpression of these fu-sion proteins up-regulated apoptosis-promoting molecules while down-regulated apoptosis-inhibiting molecules (Figure 5C) These results suggest that the RBPmotif of FHL1C is sufficient to induce apoptosis of Jurkat cells

Figure 4 FHL1C induced apoptosis of Jurkat cells through repressing RBP-J (A) Jurkat cells were transiently transfected with pEGFP, pCMX-VP16-RBP-J, pEGFP-FHL1C or pEGFP-FHL1C plus pCMX-VP16-RBP-J The percentage of apoptotic (Annexin V + ) cells

in EGFP + cell population was measured 24 h after transfection (B) Constitutively active RBP-J blocked FHL1C-induced apoptosis

in Jurkat cells Jurkat cells were transiently transfected with 1 μg

of pEGFP-FHL1C alone or in combination with increasing amounts (0.2, 0.5, 1.0 μg) of pCMX-VP16-RBP-J Cell apoptosis was measured by Annexin V staining on different days after transfection The percentages

of apoptotic (Annexin V + ) cells in EGFP + cell population were shown Bars = means ± S.D (n = 3), *P < 0.05, **P < 0.01.

Figure 5 The RBP-J-binding motif was sufficient to induce apoptosis in Jurkat cells (A) Full length and differentially truncated FHL1C (Additional file 7: Figure S3A) were inserted into pEGFPC1 in frame, and were used to transfect Jurkat cells The cells were analyzed by Annexin V staining followed by FACS 48 h post-transfection The percentages of apoptotic (Annexin V+) cells in the EGFP+cell population were determined (B) Jurkat cells were transiently transfected with plasmids as in (A) The numbers of EGFP+cells were counted at different time points after transfection (C) Jurkat cells were transiently transfected with plasmids as in (A) Cells were harvested 48 h post-transfection for RNA extraction The mRNA expression levels of Hes1, Pten, Myc, p53, Bcl2, Bax, and Caspase3 were detected by qRT-PCR, with β-actin as a reference (D) The core sequences with different length of the RBP-J-binding motif in FHL1C were fused to the 3 ′ terminus of EGFP in frame, to construct plasmids expressing EGFP with RBP-J-binding motif at the C-terminus (E) EGFP containing RBP-J-binding motif inhibited NIC-mediated transactivation of RBP-J specific reporter construct HeLa cells were transfected with different plasmids as indicated, and luciferase activity in the cell lysates was examined 48 h after transfection (F) Jurkat cells were transiently transfected with plasmids as indicated The cells were analyzed by Annexin V staining followed

by FACS 48 h after the transfection The percentages of apoptotic (Annexin V+) cells in the EGFP+cell population were determined (G) Jurkat cells were transiently transfected with plasmids as indicated The numbers of GFP+cells were counted at different time points after transfection Bars = means ± S.D (n = 3), *P < 0.05, **P < 0.01.

These results raised the possibility of developing small

peptides to disrupt Notch signaling in T-ALL cells

There-fore, as the first step, we determined which sequence in

the RBPmotifof FHL1C could induce Jurkat cell apoptosis

Oligonucleotides encoding various lengths of the RBPmotif

were synthesized, fused to the C-terminus of EGFP

(Figure 5D), and then overexpressed in Jurkat cells by

transfection All constructs exhibited suppression of

Notch-mediated transcriptional activation in reporter assays, but

the construct carrying EGFP fused to the VWWPM motif

showed suppression comparable with that of full-length

FHL1C (Figure 5E) We next examined apoptosis by

annexin-V staining In the GFP+ cell population,

overex-pression of EGFP-VWWPM efficiently induced apoptosis

of Jurkat cells, although the other two fusion proteins had

similar effects (Figure 5F) Consistently, overexpression of

EGFP fused to various lengths of the RBPmotifresulted

in a reduction of the number of transfected GFP+Jurkat

cells (Figure 5G) These results suggest that a minimal

RBP-J-binding sequence composed of five amino acids

(VWWPM) is enough to induce apoptosis of T-ALL

cells

Overexpression of FHLIC inhibits downstream genes and key pathways of notch signaling in T-ALL progression

To explore whether FHL1C-mediated apoptosis of Jurkat cells is associated with attenuation of Notch signaling,

we first examined expression of the critical downstream genes of the Notch pathway involved in T-ALL progres-sion using quantitative RT-PCR and western blotting As

a result, the mRNA levels of Hes1, Hes5, and c-Myc were significantly down-regulated by FHL1C overexpres-sion (Figure 6A) The protein level of c-Myc was also reduced remarkably (Figure 6B) These data indicate that FHL1C regulates T-ALL progression by direct suppres-sion of Notch1 target gene expressuppres-sion

Furthermore, we examined the effects of FHL1C overex-pression on the activation of PI3K/AKT and NF-κB by western blotting, which are critical pathways activated by Notch1 in T-ALL [29,30] We found that overexpression of FHL1C in Jurkat cells reduced the phosphorylation of AKT (Figure 6C and D) Activation of NF-κB is closely associated with Notch1-dependent T-ALL Therefore, we examined the levels of p50, c-Rel, and IκB in the cytosolic and nuclear

Figure 6 Overexpression of FHL1C induced apoptosis of Jurkat cells involving multiple effectors and pathways (A) Jurkat cells were transfected with pEGFP or pEGFP-FHL1C by using the Nucleofection method The cells were harvested 48 h post-transfection, and the mRNA levels of Hes1, Hes5 and c-Myc were detected by real time RT-PCR, with β-actin as a reference (B) Jurkat cells were transfected as in (A) The protein level of c-Myc was determined by using Western blotting (C,D) Cell lysates were prepared from Jurkat cells transfected with pEGFP or pEGFP-FHL1C for 48 h AKT and phosphorylated AKT (pAKT) were analyzed by Western blotting (C) The relative levels of AKT and pAKT were quantified and compared, with β-actin as an internal control (D) (E-G) Jurkat cells were transfected with pEGFP or pEGFP-FHL1C by using the Nucleofection method Cells were harvested 24 h post-transfection, and the cytosolic and nuclear extracts were fractioned P50, c-Rel and I κB were determined by Western blotting (E) The relative levels of P50 (F) and c-Rel (G) were quantified and compared, with β-actin as an internal control Bars = means ± S.D (n = 3), *P < 0.05, **P < 0.01.

fractions of FHL1C-overexpressing Jurkat cells by western

blotting The results showed that the levels of p50 and

c-Rel decreased significantly in the nuclear fraction

IκB was found primarily in the cytosolic fraction and

was also decreased slightly upon FHL1C

overexpres-sion (Figure 6E–G) This data suggest that FHL1C might

down-regulate NF-κB activity by inhibiting nuclear

trans-location of p50 and c-Rel

Discussion

The identification of activating point mutations in Notch1

in more than 50% of T-ALL cases has spurred the

devel-opment of therapies targeting the Notch1 signaling

pathway for the treatment of T-ALL To date, most of

these efforts have focused on inhibiting the activity of

γ-secretase, an enzyme that is essential for Notch

re-ceptor activation Small molecule GSIs that inhibit

γ-secretase activity have been tested in clinical trials

and shown down-regulation of Notch1 target genes in

T-ALL cells [7,31] However, GSIs are not selective for

Notch1 signaling and block other Notch receptors and

physiological pathways requiring γ-secretase Indeed,

patients have developed marked fatigue and dose-limiting

gastrointestinal toxicity in clinical trials of GSIs, because

of the inhibition of Notch1 and Notch2 in intestinal crypt

progenitors and/or stem cells, resulting in premature

differentiation into goblet cells [32] However, Real et al

subsequently showed that the gut toxicity can be

ame-liorated by combinatorial therapy using GSIs and

glu-cocorticoids [12] To avoid the side effects of GSIs,

antibodies have been developed to specifically block the

Notch1 receptor [33] However, it has been

demon-strated that the hotspot region of Notch1 mutations in

T-ALL is the PEST domain located in the C-terminus

of Notch1, which leads to delayed NIC degradation and

thus prolonged Notch signaling Therefore, these

muta-tions are less sensitive to anti-Notch antibodies [30,34]

In addition, some tumor cells harboring chromosomal

translocations or other genetic aberrations might not be

suitable for antibody-mediated therapy [35] In addition to

PEST domain mutations, another region of Notch1

muta-tions in T-ALL is the NRR region including the LNR and

HD domains, in which mutations lead to ligand

hypersen-sitivity and ligand-independent activation [7] Although

anti-NRR antibodies have been developed, sustained

treat-ment with these antibodies will likely cause vascular

neoplasms [36] More recently, Roti et al demonstrated

that inhibition of SERCA (sacro/endoplasmic reticulum

Ca2+-ATPase) calcium pumps preferentially affects the

maturation and activity of mutant Notch1 receptors,

leading to enhanced clearance of the mutant Notch

pro-tein Even if SERCA can be specifically targeted, such

inhibition does not effect on T-ALL cells with activated

Myc mutations or lacking NRR region [37]

The transactivation complex NIC-RBP-J-MAML1 is critical for signaling from Notch receptors, and is thus becoming a promising therapeutic target for T-ALL at the transcription level Recently, Moellering et al showed that SAHM1 suppresses the transcriptional complexes of Notch signaling Treatment of leukemic cells with SAHM1 inhibits cell proliferation in vitro and in a Notch1-driven T-ALL mouse model without prominent gut toxicity [16] In the current study, we found that over-expression of FHL1C induced apoptosis of the Jurkat T-ALL cell line in vitro FHL1C overexpression down-regulated c-Myc expression and attenuated the PI3K/ AKT pathway and NF-κB signaling These mechanisms may be involved in the enhanced apoptosis of Jurkat cells overexpressing FHL1C (Additional file 8: Figure S4), and suggest that FHL1C may be another therapeutic target for T-ALL at the transcriptional level Moreover,

it has been shown that Pten plays an important role in negative regulation of PI3K/AKT signaling in T-ALL However, because Jurkat cells lack active Pten protein expression, it is possible that FHL1C can suppress AKT by other mechanisms such as disruption of the NICD-P56Lck -PI3K complex [30,38,39] Further studies are needed to investigate whether FHL1C can inhibit AKT activation through Pten in native T-ALL cells

FHL1 is a member of the FHL protein family that contains four-and-a-half LIM domains FHL1 family members interact with many proteins through their LIM domains, including transcription factors, enzymes, and cytoskeleton proteins These proteins play important roles in cell differentiation and cytoskeleton formation Recent studies have shown that FHL1 also has important functions in tumorigenesis and cancer progression FHL1 expression is suppressed in a variety of tumors including lung cancer, breast cancer, brain tumors, and gastric cancer [40,41] In contrast, some reports show that FHL1 is expressed at a high level in a squamous cell carcinoma cell line [42] FHL1 is aberrantly expressed in most T-ALL cell lines, particularly those exhibiting deregu-lated TLX1/HOX11 expression after specific chromosome translocation [43] In our study using PBMCs from T-ALL patients, we detected FHL1A expression in two cases, but the significance and underlying mechanism are unclear We also detected significant down-regulation of FHL1C expression in PBMCs of T-ALL patient, accom-panied by up-regulation of Hes1, a Notch target gene involved in T-ALL progression These results suggest that FHL1C may be involved in T-ALL progression and can be used as a therapeutic target of the disease However, the mechanism regulating FHL1C expression in T-ALL cells remains unknown, and whether FHL1C is involved in other cancers is unclear In addition, although FHL1B (KyoT3) is another isoform of FHL1, which encodes a

34 kDa polypeptide containing the same RBPmotiffound in